Download Exam 1 PS 217, Spring 2010 Convert to z

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Inductive probability wikipedia , lookup

Foundations of statistics wikipedia , lookup

Bootstrapping (statistics) wikipedia , lookup

History of statistics wikipedia , lookup

Taylor's law wikipedia , lookup

Gibbs sampling wikipedia , lookup

Misuse of statistics wikipedia , lookup

Resampling (statistics) wikipedia , lookup

Student's t-test wikipedia , lookup

Transcript
Exam 1
PS 217, Spring 2010
1. Answer the following questions assuming that they are dealing with a population of SAT-M scores, which are
normally distributed with µ = 500 and σ = 100.
a. What is the probability that a person would achieve SAT-M scores between 550 and 650? [3 pts]
Convert to z-score Determine Probability
550 − 500
.1915
z=
= .5
100
650 − 500
.4332
z=
= 1.5
100
€
€
€
€
Subtract smaller probability from the larger
probability to yield answer => .2417.
b. What is the probability that a person would achieve SAT-M scores between 430 and 480? [3 pts]
Convert to z-score
430 − 500
z=
= −.7
100
480 − 500
z=
= −.2
100
Determine Probability
.2580
Subtract smaller probability from the larger
probability to yield answer => .1787.
.0793
c. What SAT-M scores would be achieved by the lower 65% of the population? [3 pts]
Look up .65 in the body of the distribution
(Column B), then find the z-score (Column
A), which is .385. Then, use the z-score
formula to determine the SAT-M score =>
538.5.
X − 500
, X = 538.5
100
.385 =
€
€
d. What SAT-M scores would be achieved by the upper 5% of the population? [3 pts]
Look up .05 in the tail of the distribution
(Column C), then find the z-score (Column
A), which is 1.645. Then, use the z-score
formula to determine the SAT-M score =>
664.5.
1.645 =
X − 500
, X = 664.5
100
€
€
e. What is the probability that a sample of n = 25 would yield a mean (M) SAT-M of 440 or less from this
population? [4 pts]
Convert the M = 440 to a z-score, which
means that you must first compute the
standard error (20). Look up the probability
for z = -3 in the tail of the distribution
(Column C) => .0013.
100
= 20
25
440 − 500
z=
= −3
20
σX =
€
f. For samples of n = 100, what mean SAT-M scores would comprise the middle 90% of the sampling distribution of
the mean? [4 pts]
€
Look up .05 in the tail (Column C) to find
the corresponding z-score => 1.645. Then,
determine the standard error (10) and use
the z-score formula to determine the means.
€
€
100
X − 500
= 10 1.645 =
, X = 516.45
10
100
X − 500
, X = 483.55
−1.645 =
10
€
€
σX =
Page 1 of 4
€
2a. Below is a sample of quiz scores. Estimate µ, σ2, and σ of the population from which the sample was drawn. [5
pts]
2
Quiz
107
Quiz
X = M = µˆ =
= 7.13
8
64
15
7
49
107 2
3
9
SS = 819 −
= 55.7
8
64
15
9
81
55.7
s2 = σˆ 2 =
= 3.98
10
100
14
7
49
s = σˆ = 1.995
6
36
8
64
9
81
7
49
6
36
€
8
64
3
9
8
64
Sum
107
819
2b. If you were testing H0: µ = 8.5, what statistic would you use? Why? [2 pts]
You’d use a t-test, because σ isn’t known.
2c. If you were testing H0: µ = 8.5, what would be your effect size d? [3 pts]
d=
€
7.13 − 8.5
= .69
1.995
3a. As you know, gestation periods are normally distributed with µ = 268 and σ = 16. Imagine that you are giving a
talk to 64 pregnant women who happen to be cigarette smokers. Suppose that you want to tell them what the
average (mean) gestation period would be for a group that size and you want to be 90% confident in your estimate.
What would you tell them about the upper and lower limits that would encompass the typical mean gestation period
for a group of 64 women? (In other words, what gestation periods cut off the upper and lower 5% of the
distribution?) [5 pts]
Look up .05 in the tail (Column C) to find
the corresponding z-score => 1.645. Thus,
±1.645 are the z-scores that cut off the upper
and lower 5%. Compute σ M and then use
the z-score formula to determine the lower
€
(264.7) and upper (271.3) values.
16
X − 268
= 2 1.645 =
, X = 271.3
2
64
X − 268
, X = 264.7
−1.645 =
2
€
€
σX =
€
3b. Suppose that for this sample of 64 smokers,€
the mean gestation period M = 260 days. How likely is it that your
group was sampled from the normal distribution described in 3a?
State the null and alternative hypotheses you would be using to test the assertion. [1 pt]
H0: µ = 268
H1: µ ≠ 268
Page 2 of 4
Test the null hypothesis. What you would conclude? (Be explicit!!) [9 pts]
Decision Rule: If |z| ≥ 1.96, then reject H0.
z=
€
260 − 268
= −4
2
Decision: Reject H0, because |-4| ≥ 1.96.
Conclude: Sample likely came from a population with µ < 268.
Tell me, in words, what kind of error you might be making in your conclusion. [2 pts]
Because you’re rejecting H0, you could be making a Type I error (rejecting H0 when H0 is
true).
4. On the curves seen below, label the areas that represent Type I Errors, Type II Errors, Power, and Correct
“Retention.” [5 pts]
5. What is the general formula for all standard scores? [2 pts]
Score of Interest - Mean of Distribution
Standard Deviation of Distribution
€
6. Suppose that you take one sample of n = 16 and another sample of n = 100. What does the Central Limit Theorem
tell you about the shape and variability of the sampling distribution of the mean from which each sample mean
would have been drawn? [5 pts]
Mean of the sampling distribution would be µ. However, the variability of the sampling
distribution of the mean would decrease as the sample size increases. Thus, with n = 16, the
standard error would be the population standard deviation divided by 4 and with n = 100
the standard error would be the population standard deviation divided by 10. Moreover, as
the sample size increases, the sampling distribution of the mean would become increasingly
normal in shape.
Page 3 of 4
7. What is the (small) advantage of using the standard deviation instead of the variance as a measure of variability?
[2 pts]
The standard deviation is expressed in the same units of measurement of the scores (e.g.,
inches) instead of squared scores (e.g., inches squared).
8. Under which conditions might the median be a better measure of central tendency than the mean? [2 pt.]
When the distribution is skewed.
9. Interpret the PASW/SPSS output below as completely as you can. (In other words, what is the statistic, what are
the scores, how would you interpret the results, etc.) [3 pts]
It’s a single-sample t-test on quiz scores. H0: µ = 8.5 and H1: µ ≠ 8.5. With t(14) = -2.653, p
= .019. Because p ≤ .05, you would reject H0 and conclude that the sample was drawn from
a population with µ < 8.5.
10. Adding a constant (e.g., 3) to a set of scores has what effect on the mean and standard deviation of the
transformed scores? [2 pts]
The mean would increase by that constant and the standard deviation would be
unchanged.
11. Transforming a set of scores into z-scores has what effect on the mean and standard deviation of the transformed
scores? [2 pts]
The mean would be 0 and the standard deviation would be 1.
12. Signal Detection Theory shares many characteristics with Null Hypothesis Significance Testing (NHST). [2 pts]
What would be the Signal Detection analog of effect size?
What would be the NHST analog of proportion hits?
d’
power
13. The probability of making a Type I Error is typically set to 5%. [1 pt]
14. As the sample size approaches infinity, what happens to the value of tCritical and why? [2 pts]
Because the distribution of t would become more normal, tCritical would become 1.96.
15. In a positively skewed distribution, what is the relationship between the mode, median, and mean? [2 pts]
Mean > Median > Mode
Page 4 of 4